The new age of bioimaging — Page 2 of 7 <
Back Next
>
Faces of
mitosis
Interest in facial recognition software increased in
recent years. Computers can “recognize”
individuals’ faces, as they can identify individuals’
fingerprints. Postdoctoral researcher Jason Moffat of
David
Sabatini’s lab employs similar technology
to identify genes that could be involved in mitosis
(cell division).
“In Jason’s experiment, programs sorted
pictures or ‘portraits’ of cells by examining ‘features’ such as cell shape and DNA distribution,”
says Sabatini.
| “The literature on mitosis in mammals is
still murky,” says Moffat. “We’re
fairly confident that other labs will eventually
confirm the role of these genes.” |
 |
Jason Moffat
|
Moffat wanted to apply image-analysis techniques to
identify genes required for cell growth and division.
He also wanted to show scientists how to use a new tool
developed by a team at Whitehead and the Broad Institute.
The collaborators manufactured special viruses, designing
them to infect cells and silence specific human and
mouse genes. The team built thousands of unique viruses,
and Moffat decided to use the resulting library to perform
a massive gene-knockout experiment.
After weighing several potential projects, he chose
to focus on mitosis. He planned to screen more than
1,000 human genes to determine which ones play a role
in the cell division process. Scientists had already
identified some of the genes involved, giving him a
basis for comparison.
Next, Moffat turned to a Cellomics ArrayScan automated
fluorescence microscopic-imaging system designed for
high-content screening. He prepared hundreds of plates,
placing human cells in thousands of wells. Each well
received a different virus. After waiting for the vectors
to work, the microscope took pictures of each well.
But what did those pictures tell Moffat about cell division?
After allowing the viruses to knock down their intended
target genes, he added fluorescent labels for three
components of the cells—DNA, actin (to outline
the cell shape) and a structural protein that is detectably
modified throughout mitosis. Thus the pictures allowed
him to flag cells with abnormal mitotic activity. He
relied on two software packages, one being the CellProfiler
created by Anne Carpenter, to expedite the flagging
process. The programs analyzed the visual properties
of each cell, comparing them to the properties of a
“normal” cell.
The project implicated dozens of genes in mitosis,
though previous studies confirm only a handful of those
links. Cell published the results in March.
“The literature on mitosis in mammals is still
murky,” says Moffat. “We’re fairly
confident that other labs will eventually confirm the
role of these genes.”
| Written by Alyssa Kneller |
|
